Intel Stops The Tick-Tock Clock

[jasonf2]

We just said the same thing. Only what you are terming as price gouging is what I am calling profit margin. Those profit margins in the desktop space have also funded the process node R&D. Another fabless ARM licensed company mediatek is exactly what I am talking about here.

Fabless companies do not get fab services for free. The fees they pay to have those wafers made by someone else cover that someone else's costs to get the process R&D done and span everything from SRAM chips to SoCs and GPUs, not just CPUs. Everything being made by a foundry pays for some fraction of the R&D. In the case of Samsung and nearly everyone else other than Intel, they don't even make desktop or server chips for themselves, the only profit they see "from desktop and server" is the margin they make on someone else's wafer starts.

I never said that fabless companies get their product made for free. Globalfoundries, Samsung, TSMC, UMC and Intel (on a different scale) all get paid for their services. The difference between Intel and MediaTek (for example) is that Mediatek only handles part of the design of the SOC. They choose who they would like to make it for them (usually determined by price and performance metric of the fab process nodes available) and have some customization necessary for production process and tweak the ARM baseline designs. Intel on the other hand handles the whole thing. They design the chip, design the fabrication node, build the fabrication node and actually build the chip. They have been able to keep doing this by having a good profit margin in their products. Consumers in the desktop market have been willing to pay because no other manufacture/designer has really been able to close in on the performance lead that Intel has had since about the I7-920 (Nehalem) came out in response to the Athalon64 (I know I am going tangent here but the 920 was the real game changer not Sandybridge.) . This has not been the case in the mobile space.
Qualcomm and Apple (also fabless) have been to the mobile space what Intel has been to desktop. The manufactures in the mobile space (except Apple) have been running on razor thin margins from the very beginning. Samsung as a great example has used Snapdragon processors in their flagship phones for a long time. But when their in-house designs (Samsung's) are better than what Qualcomm is putting out they kick them to the curb (just look at processors installed in generations of the Galaxy Note lineup). If Qualcomm can make up the tech in the next generation they might be back in. Intel in my opinion doesn't have a mobile space product available with enough of a performance lead to justify manufactures to put them in a flagship device or a cheap enough one to put them in a giveaway device. And the Atom trademark name has a stigma of being a "good enough" product not a flagship lineup (remember netbooks). Intel needs that premium in their current production model for their business to work. Playing arm chair football here (which we all are) I would say that what Intel needs to do to get into the mobile game is come up with a mobile chip powerhouse with a 50%+ all metrics advantage and establish a premium trademark on the product. I don't really have any illusions of that being reality with $40 per chip being the top end of the mobile game. What I see to be more likely is that the "Core" lineup moves down in power consumption enough to get manufactures to start packaging it into ultra premium devices like a future Surface Phone which in essence does the same thing. If they can keep their 30% power reduction trend going they are well within reach of this on a 3000-4000 mah battery device in Cannonlake which will take a little longer now, but still in the pipe. Until phones are being used as workstation docks and able to run legacy windows software there really isn't even a market here yet though. So a lot hinges on what happens for the surface phone for Wintel and that Atom chip they are putting in there. The biggest reason here though isn't the device itself though, it is the price of the device. Intel needs these things to cost $1200 and up for the premium margin SOC to work. This may sound a little high for a phone, but I would remind you what happened in the tablet segment as soon as Microsoft brought the Surface Pro lineup in. I really look forward to purchasing my Surface phone pro with an I5 (gen 8 from the looks of it now) in it and hooking it to my house/work dock with 802.11ad (60ghz) charging on a wireless charger pad while playing chrysalis (Just had to get that in there.) and complaining that gen 8 only had a 10% performance increase with a 30% reduction in power consumption. But I have no illusions that I am going to do that on a current flagship phone price in the near future.

 

[InvalidError]

Consumers in the desktop market have been willing to pay because no other manufacture/designer has really been able to close in on the performance lead that Intel has had

No manufacturer is legally capable of competing against Intel's x86 and related extensions because all designers with x86 ISA licenses apart from AMD have bailed out of the x86 PC/server chip business to reduce or eliminate their licensing burden and Intel is not licensing the x86 ISA to new comers. There is a considerable difference between people being "willing to pay" and people having practically no other choice since Intel is currently in a monopoly position for x86 chips due to lack of effective competition.

 

[eltoro]

Intel is not immune to the laws of physics and the increasing difficulties that come along with going smaller.

If going forward was so easy, the other chip designers and makers would have no trouble catching up with Intel. But they are all struggling just as much. Like it or not, that's what happens when you get close to physical limits.

The only practical way to increase performance much beyond the current state of things is to add more CPU cores but there is too little mainstream software capable of leveraging multi-core CPUs in a meaningful and efficient way for heavily multi-core/multi-thread mainstream CPUs to make sense.

The main difference between Intel and its competitors is their financial and infrastructure strength. When Intel feels AMD at its heels, it pumps more money into the development. But now there's no point in spending all that money when they could wait at least a couple of years for AMD to catch up. That's what I mean when I say that Intel is allowing itself to run into difficulties. Solving them faster costs more money than a slower development cycle.

 

[MisterZ]

Still no reason to upgrade from a Quad core Q6600. Wake me up when an interesting new CPU comes out.

 

[jasonf2]

Consumers in the desktop market have been willing to pay because no other manufacture/designer has really been able to close in on the performance lead that Intel has had

No manufacturer is legally capable of competing against Intel's x86 and related extensions because all designers with x86 ISA licenses apart from AMD have bailed out of the x86 PC/server chip business to reduce or eliminate their licensing burden and Intel is not licensing the x86 ISA to new comers. There is a considerable difference between people being "willing to pay" and people having practically no other choice since Intel is currently in a monopoly position for x86 chips due to lack of effective competition.

I am pretty sure Via is still in there (on the desktop side) there are still quite a few on the SOC side that are in license. My understanding is though the license isn't just a one way road either. It requires that no two manufactures can build on the same motherboard socket (So AMD cant just plug in their processor on an Intel Chipset or vice versa.) and there is from what I can tell a Patent sharing agreement that is non transferable or able to be canceled at will (Intel uses a fair amount of AMD IP in their products). Also there is a list of Fabs licensed to produce X86, not just anyone can build the chip. With all of those hoops and the obviously massive expense of not only having to design the chip, a chip set and motherboard in desktop I doubt that they have too many companies lined up to sign on either. Especially in a shrinking market. SOC on the other hand has had some new companies in the last couple of years sign on (or at least speculatively in the works). Over the last couple of decades though there have been at least half a dozen chip producers compete in the desktop space and all but AMD have been unable to gain traction and failed. Most were assumed ultimately by AMD or Intel for their IP. AMD is still surviving, but routinely looses money. So while I would put anyone in the desktop market against Intel as a David and Goliath comparison (with Goliath wearing a steel helmet) monopoly is really pretty harsh.

 

[Andress Constantinides]

Does this points to the fact that Intel is the stronger believer of multitasking computers will never age, until their technology seize to function?

 

[jimmysmitty]

Invalid Your price points are off. Intel has been artificially under pricing its mobile platform for some time to try to make market penetration along with special vendor arrangements.

Every other mobile SoC costs from $10 for low-end Mediatek chips to $45 for the Nvidia's X1 and other high-end chips. If Intel cannot produce competitive mobile chips priced within that price range, then they are doomed as far as mobile is concerned.

If you think my price points are "way off", consider this: you can get 8-10" tablets based on Mediatek SoCs for under $100 and 7" ones for as little as $50. There is very little room in the mobile SoC space for Intel to indulge in the same massive price gouging they have been doing in laptops and desktops.

We just said the same thing. Only what you are terming as price gouging is what I am calling profit margin. Those profit margins in the desktop space have also funded the process node R&D. Another fabless ARM licensed company mediatek is exactly what I am talking about here.

I would rather say their margins in the server market funded their R&D. While desktop margins are good, their server market margins are vastly superior and right now they do not have much if any competition from AMD in that market since, well BD just is not efficient enough.

Still no reason to upgrade from a Quad core Q6600. Wake me up when an interesting new CPU comes out.

That is vastly untrue. I had a Q6600 and even overclocked that thing was slower than a stock i5 2500K. It was a great chip no doubt but there is also just too much around the CPU that has changed and gotten faster that makes the Q6600 a outdated chip.

 

[InvalidError]

I am pretty sure Via is still in there (on the desktop side) there are still quite a few on the SOC side that are in license.

While Intel still has some outstanding x86 licenses, none of them are active much beyond niche markets. Most former licensees simply sold their designs to other licensees and bailed out.

If you go check Via's website, you will see that Via has not launched or announced any new CPUs since 2011 and the current C7/Eden/Nano, which are 40nm parts, are only an update of their 2008 design to add things like SSE4 to some models.

There was a demo of "Isaiah II" in 2014 which supposedly outperformed Intel's Baytrail in benchmarks and was meant to launch in late 2014 but there has been no signs of it since.

Via's newest currently shipping x86 chip is practically eight years old and if the paper-launched newer version is only on par with Intel's mobile SoC from two generations ago, the much older previous design should be slower. Not what I would consider a threat to Intel and AMD's desktop and laptop CPUs.

That's 5-8 years between product updates. Not something you see in PCs and mobile devices. There is one market where 5-8 years product cycles make sense: embedded where the appliance is intended to see 5+ years of service life and the end-user is not expected to care what CPU or OS the application is running on as long as it works as expected.

 

[jasonf2]

I am pretty sure Via is still in there (on the desktop side) there are still quite a few on the SOC side that are in license.

While Intel still has some outstanding x86 licenses, none of them are active much beyond niche markets. Most former licensees simply sold their designs to other licensees and bailed out.

If you go check Via's website, you will see that Via has not launched or announced any new CPUs since 2011 and the current C7/Eden/Nano, which are 40nm parts, are only an update of their 2008 design to add things like SSE4 to some models.

There was a demo of "Isaiah II" in 2014 which supposedly outperformed Intel's Baytrail in benchmarks and was meant to launch in late 2014 but there has been no signs of it since.

Via's newest currently shipping x86 chip is practically eight years old and if the paper-launched newer version is only on par with Intel's mobile SoC from two generations ago, the much older previous design should be slower. Not what I would consider a threat to Intel and AMD's desktop and laptop CPUs.

That's 5-8 years between product updates. Not something you see in PCs and mobile devices. There is one market where 5-8 years product cycles make sense: embedded where the appliance is intended to see 5+ years of service life and the end-user is not expected to care what CPU or OS the application is running on as long as it works as expected.

I don't disagree that there has been little activity in the marketplace other than AMD and Intel. But I would still solidly defend that this isn't some licensing conspiracy from Intel. Intel's solid foundry capabilities, design capabilities and cash war-chest coupled with AMD's mainstream pricing makes the X86 desktop market a very uninviting place for big money investment in a X86 startup to really work. When I say big money I don't know how many billions it would take to really make a market ready chip let alone sell the thing effectivly. That even being fabless. The last one that I thought had a real shot was Transmeta. Back when everything X86 didn't really care about power they had a laptop chip that was pretty darn efficient at their fab node. They even had some big Japanese manufactures on board. All it took for them to break though was for a couple of QC issues, some marketing / price shift from Intel and they were gone before their pre-ship enigmatic web page was. They all start an IPO, the investors expect return and when it doesn't happen as quickly as they expect the rug gets pulled out from under them.

This is just a really tough market and Intel is so good at playing the game here no one wants to play with them anymore.

 

[jimmysmitty]

I am pretty sure Via is still in there (on the desktop side) there are still quite a few on the SOC side that are in license.

While Intel still has some outstanding x86 licenses, none of them are active much beyond niche markets. Most former licensees simply sold their designs to other licensees and bailed out.

If you go check Via's website, you will see that Via has not launched or announced any new CPUs since 2011 and the current C7/Eden/Nano, which are 40nm parts, are only an update of their 2008 design to add things like SSE4 to some models.

There was a demo of "Isaiah II" in 2014 which supposedly outperformed Intel's Baytrail in benchmarks and was meant to launch in late 2014 but there has been no signs of it since.

Via's newest currently shipping x86 chip is practically eight years old and if the paper-launched newer version is only on par with Intel's mobile SoC from two generations ago, the much older previous design should be slower. Not what I would consider a threat to Intel and AMD's desktop and laptop CPUs.

That's 5-8 years between product updates. Not something you see in PCs and mobile devices. There is one market where 5-8 years product cycles make sense: embedded where the appliance is intended to see 5+ years of service life and the end-user is not expected to care what CPU or OS the application is running on as long as it works as expected.

I don't disagree that there has been little activity in the marketplace other than AMD and Intel. But I would still solidly defend that this isn't some licensing conspiracy from Intel. Intel's solid foundry capabilities, design capabilities and cash war-chest coupled with AMD's mainstream pricing makes the X86 desktop market a very uninviting place for big money investment in a X86 startup to really work. When I say big money I don't know how many billions it would take to really make a market ready chip let alone sell the thing effectivly. That even being fabless. The last one that I thought had a real shot was Transmeta. Back when everything X86 didn't really care about power they had a laptop chip that was pretty darn efficient at their fab node. They even had some big Japanese manufactures on board. All it took for them to break though was for a couple of QC issues, some marketing / price shift from Intel and they were gone before their pre-ship enigmatic web page was. They all start an IPO, the investors expect return and when it doesn't happen as quickly as they expect the rug gets pulled out from under them.

This is just a really tough market and Intel is so good at playing the game here no one wants to play with them anymore.

Considering that AMD dropped their FABs to save money on the R&D it would cost quite a bit to make a competitive x86 CPU that can match Intel, especially with Intels Fabrication advantage. Even Samsung's 14nm isn't a match for Intels yet, and I doubt GloFlos will be since Samsung has more experience in Fabrication than GloFlo does.

Most of these Intel competitive process nodes are coming from consortium's now though as no single company has the experience or financial capital to compete with Intel on the same level. As I have said before FAB42 was $5 billion alone, more than AMDs net worth. Hell they paid cash for Altera, $16.7billion. Thats some major money that no other competitor has, even Samsung who has great devices in multiple markets and has a lot of hands in DRAM/NAND.

 

[InvalidError]

I don't disagree that there has been little activity in the marketplace other than AMD and Intel. But I would still solidly defend that this isn't some licensing conspiracy from Intel.

Intel has sued just about every x86 ISA licensee in the past to get them to drop their licenses. That's on top of questionable business practices, price dumping to run competition out of the market, strong-arming computer vendors to stop them from offering competing products, etc.

While Intel has gone quiet on the license dispute side, there was a time where they were going all-out to make their licensees' lives miserable. A bunch of them folded, most of the rest either merged, sold out or backed out of the PC/server market.

 

[RedJaron]

They should call the new development scheme "jump, jive, and wail".

I'll call it Tick, Tock, Tuck. Tuck things away that aren't needed, just finish things off and refine them.

 

[jasonf2]

I don't disagree that there has been little activity in the marketplace other than AMD and Intel. But I would still solidly defend that this isn't some licensing conspiracy from Intel.

Intel has sued just about every x86 ISA licensee in the past to get them to drop their licenses. That's on top of questionable business practices, price dumping to run competition out of the market, strong-arming computer vendors to stop them from offering competing products, etc.

While Intel has gone quiet on the license dispute side, there was a time where they were going all-out to make their licensees' lives miserable. A bunch of them folded, most of the rest either merged, sold out or backed out of the PC/server market.

Oracle vs Google
Samsung vs Apple
Intel vs Nvidia
Nvidia vs Samsung

just a few going on right now

Patent trolling is and has been part of this business since it started. I don't really know if there is a single tech company out there that doesn't have some sort of an active patent infringement lawsuit on them.

This is obviously going nowhere. Intel is the devil. But by the rules laid out here every market dominant company in the history of time is also the devil.

You win.

 

[turkey3_scratch]

I don't disagree that there has been little activity in the marketplace other than AMD and Intel. But I would still solidly defend that this isn't some licensing conspiracy from Intel.

Intel has sued just about every x86 ISA licensee in the past to get them to drop their licenses. That's on top of questionable business practices, price dumping to run competition out of the market, strong-arming computer vendors to stop them from offering competing products, etc.

While Intel has gone quiet on the license dispute side, there was a time where they were going all-out to make their licensees' lives miserable. A bunch of them folded, most of the rest either merged, sold out or backed out of the PC/server market.

I'm a strong believer that when it comes to business, as long as it's a legal practice, there is nothing wrong with a company like Intel trying to run those others out of the market.

 

[InvalidError]

Patent trolling is and has been part of this business since it started.

There is a "small" difference between patent trolling and unilaterally terminating licenses.

AMD originally started making x86 chips because Intel did not have enough fab capacity at the time and had to hire AMD for extra capacity. When Intel caught up with demand, they wanted to rescind AMD's x86 license but the DoJ stepped in to prevent Intel from establishing a monopoly for the x86 chips most of the government depends on.

Were it not for government action, AMD as an x86 designer would have been history some 20 years ago.

 

[sykozis]

Consumers in the desktop market have been willing to pay because no other manufacture/designer has really been able to close in on the performance lead that Intel has had

No manufacturer is legally capable of competing against Intel's x86 and related extensions because all designers with x86 ISA licenses apart from AMD have bailed out of the x86 PC/server chip business to reduce or eliminate their licensing burden and Intel is not licensing the x86 ISA to new comers. There is a considerable difference between people being "willing to pay" and people having practically no other choice since Intel is currently in a monopoly position for x86 chips due to lack of effective competition.

I am pretty sure Via is still in there (on the desktop side) there are still quite a few on the SOC side that are in license. My understanding is though the license isn't just a one way road either. It requires that no two manufactures can build on the same motherboard socket (So AMD cant just plug in their processor on an Intel Chipset or vice versa.) and there is from what I can tell a Patent sharing agreement that is non transferable or able to be canceled at will (Intel uses a fair amount of AMD IP in their products). Also there is a list of Fabs licensed to produce X86, not just anyone can build the chip. With all of those hoops and the obviously massive expense of not only having to design the chip, a chip set and motherboard in desktop I doubt that they have too many companies lined up to sign on either. Especially in a shrinking market. SOC on the other hand has had some new companies in the last couple of years sign on (or at least speculatively in the works). Over the last couple of decades though there have been at least half a dozen chip producers compete in the desktop space and all but AMD have been unable to gain traction and failed. Most were assumed ultimately by AMD or Intel for their IP. AMD is still surviving, but routinely looses money. So while I would put anyone in the desktop market against Intel as a David and Goliath comparison (with Goliath wearing a steel helmet) monopoly is really pretty harsh.

The only companies in the last 2 decades to attempt to compete with Intel directly in the consumer market have been AMD and Cyrix. Of those 2, AMD is the only one left. Cyrix was bought out by VIA, who already had an x86 license of their own. VIA hasn't released a desktop processor since the C7 in 2007. VIA has chosen to cater to the embedded market exclusively as it's the only remaining x86 market that Intel doesn't dominate. Some VIA chips have found their way into low-end tablets, but VIA is incapable of competing with ARM in the tablet market.